The goal of this project is to understand the role that conduct disorder plays in the development of substance-use problems in young adulthood. To reach this goal, the proposed project will amplify two ongoing, federally funded, multi-site, longitudinal studies. The Child Development Project (CDP) is a prospective study of 585 youth (50% female; 19% minority) from three sites, who have been followed since age four years and are now 19 years old (with 84% retention of original sample). The Fast Track (FT) Project is a prevention experiment in which 892 early-starting conduct-problem youth (31% female; 55% minority) from four sites were identified in first grade and randomly assigned to receive a ten-year intervention or not and have been followed through age 18 (83% retention of sample). Findings indicate that youth assigned to the FT intervention display 40% fewer diagnoses of conduct disorder in high school than controls. FT also includes a normative-representative sample of 387 youth (50% female; 49% minority) who are being followed to study how conduct disorder develops. To date, substance use in both the CDP and FT has been measured but not in depth. For both studies, new data collection for the proposed project will include detailed measurement of participants' use, dependence, and abuse of illicit substances, alcohol, and tobacco in early adulthood. Three aims will be addressed. The first aim is to understand how early conduct disorder leads to substance-use problems. A transactional-developmental model will be tested that posits risk and protective factors in the domains of child behavior, socio-cultural contexts, parenting, peer relations, schooling, and social cognitions. Findings that emerge from the CDP will be replicated with the FT normative sample. The second aim is to understand resilience to substance-use development among conduct-problem children. The FT high-risk control group, known to be at elevated risk for substance use, will be examined to identify protective factors that account for resilience to substance use. The hypothesis will be tested that professional services moderate risk; specifically, services that aggregate deviant children in groups will enhance risk, whereas individual services that focus on family process will decrease risk. The third aim is to test the efficacy of the FT intervention in preventing substance-use problems in early adulthood and the role of reduced conduct disorder as a mediator of this effect.